[00/12] target/arm: VFP decodetree conversion followups

Message ID 20190613163917.28589-1-peter.maydell@linaro.org
Headers show
Series
  • target/arm: VFP decodetree conversion followups
Related show

Message

Peter Maydell June 13, 2019, 4:39 p.m.
This patchset does a couple of the cleanup/leftover things noted
in the coverletter of the vfp decodetree conversion or discussed
in code review:
 * use vfp_expand_imm() for doing the VFP const-immediate
   decode, rather than hand-coding it
 * get rid of the final uses of cpu_F[01][sd]

cpu_V0, cpu_V1 and cpu_M0 will have to wait until somebody
cares to tackle the Neon decode logic and the iwmmxt decode...

thanks
-- PMM

Peter Maydell (12):
  target/arm: Move vfp_expand_imm() to translate.[ch]
  target/arm: Use vfp_expand_imm() for AArch32 VFP VMOV_imm
  target/arm: Stop using cpu_F0s for NEON_2RM_VABS_F
  target/arm: Stop using cpu_F0s for NEON_2RM_VNEG_F
  target/arm: Stop using cpu_F0s for NEON_2RM_VRINT*
  target/arm: Stop using cpu_F0s for NEON_2RM_VCVT[ANPM][US]
  target/arm: Stop using cpu_F0s for NEON_2RM_VRECPE_F and
    NEON_2RM_VRSQRTE_F
  target/arm: Stop using cpu_F0s for Neon f32/s32 VCVT
  target/arm: Stop using cpu_F0s in Neon VCVT fixed-point ops
  target/arm: stop using deprecated functions in NEON_2RM_VCVT_F16_F32
  target/arm: Stop using deprecated functions in NEON_2RM_VCVT_F32_F16
  target/arm: Remove unused cpu_F0s, cpu_F0d, cpu_F1s, cpu_F1d

 target/arm/translate-a64.h     |   1 -
 target/arm/translate.h         |   7 +
 target/arm/translate-a64.c     |  32 -----
 target/arm/translate-vfp.inc.c |  61 +++++----
 target/arm/translate.c         | 240 ++++++++++++---------------------
 target/arm/vfp.decode          |  10 +-
 6 files changed, 133 insertions(+), 218 deletions(-)

-- 
2.20.1

Comments

Philippe Mathieu-Daudé June 14, 2019, 5:27 a.m. | #1
On 6/13/19 6:39 PM, Peter Maydell wrote:
> This patchset does a couple of the cleanup/leftover things noted

> in the coverletter of the vfp decodetree conversion or discussed

> in code review:

>  * use vfp_expand_imm() for doing the VFP const-immediate

>    decode, rather than hand-coding it

>  * get rid of the final uses of cpu_F[01][sd]

> 

> cpu_V0, cpu_V1 and cpu_M0 will have to wait until somebody

> cares to tackle the Neon decode logic and the iwmmxt decode...

> 

> thanks

> -- PMM

> 

> Peter Maydell (12):

>   target/arm: Move vfp_expand_imm() to translate.[ch]

>   target/arm: Use vfp_expand_imm() for AArch32 VFP VMOV_imm

>   target/arm: Stop using cpu_F0s for NEON_2RM_VABS_F

>   target/arm: Stop using cpu_F0s for NEON_2RM_VNEG_F

>   target/arm: Stop using cpu_F0s for NEON_2RM_VRINT*

>   target/arm: Stop using cpu_F0s for NEON_2RM_VCVT[ANPM][US]

>   target/arm: Stop using cpu_F0s for NEON_2RM_VRECPE_F and

>     NEON_2RM_VRSQRTE_F

>   target/arm: Stop using cpu_F0s for Neon f32/s32 VCVT

>   target/arm: Stop using cpu_F0s in Neon VCVT fixed-point ops

>   target/arm: stop using deprecated functions in NEON_2RM_VCVT_F16_F32

>   target/arm: Stop using deprecated functions in NEON_2RM_VCVT_F32_F16

>   target/arm: Remove unused cpu_F0s, cpu_F0d, cpu_F1s, cpu_F1d

> 

>  target/arm/translate-a64.h     |   1 -

>  target/arm/translate.h         |   7 +

>  target/arm/translate-a64.c     |  32 -----

>  target/arm/translate-vfp.inc.c |  61 +++++----

>  target/arm/translate.c         | 240 ++++++++++++---------------------

>  target/arm/vfp.decode          |  10 +-

>  6 files changed, 133 insertions(+), 218 deletions(-)

> 


Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>